[Artificial intelligence-guided precision medicine in hematological disorders]

Area of Science:

• Oncology
• Genomics
• Medical Informatics
• Artificial Intelligence

Background:

• Precision medicine in oncology relies on genomic data for targeted patient interventions.
• Next-generation sequencing (NGS) is crucial, but manual interpretation of somatic mutations is labor-intensive.
• Integrating clinical and genomic data for hematological malignancies requires efficient data curation.

Purpose of the Study:

• To introduce the manual interpretation process in clinical sequencing (CS) for a non-specialist audience.
• To provide an overview of an artificial intelligence (AI) platform, Watson for Genomics.
• To highlight the necessity and limitations of AI in CS data interpretation for hematological cancers.

Main Methods:

• Formation of a clinical sequencing (CS) team since 2015 to integrate clinical and genomic information.
• Utilizing artificial intelligence (AI) to assist in the collation of CS data.
• Analysis of AI-assisted CS data from over 300 patients with hematological cancers.

Main Results:

• AI significantly aids in the manual interpretation process of complex genomic and clinical data.
• AI-assisted CS data has been successfully collated for a substantial cohort of hematological cancer patients.
• The study identified specific pitfalls and limitations of AI in interpreting CS data for hematologists.

Conclusions:

• AI is essential for improving the efficiency and accuracy of manual interpretation in clinical sequencing.
• Hematologists must be aware of AI's capabilities and limitations when using AI-generated outputs for clinical decisions.
• Continued development in medical informatics and AI is crucial for advancing precision medicine in oncology.